Don't Show Again Yes, I would!

ChatGPT 4: Everything we know so far

ChatGPT 4 is coming, and rumor has it that it could bring massive improvements to the already impressive language skills in OpenAI’s ChatGPT.

To be clear, ChatGPT 4 isn’t likely to be OpenAI’s next product name, but we took a little creative license and combined the ChatGPT name with the improved AI model that will drive it in the future, GPT-4. Let’s dive into GPT-4, how ChatGPT works now, and when OpenAI might release its next major upgrade.

What is GPT-4?

Midjourney's cartoon of a robot talking to a woman using a laptop.
A Midjourney cartoon where a robot talks to a woman using a laptop. The introduction of Midjourney is really motivated by Tracy

GPT-4 is a new language model created by OpenAI that can generate text similar to human speech. It will develop the technology used by ChatGPT, which is based on GPT-3.5. GPT stands for Generative Pre-trainer Transformer, which is a deep learning technology that uses artificial neural networks to write like a human.

Billions of parameters are used to adjust the output of simulated neurons in ChatGPT to achieve the desired result. The sheer amount of detail helps the computer get closer to the interconnectedness of the human brain, which also contains billions of neurons. OpenAI has not confirmed any details about GPT-4 but acknowledges that it is a work in progress. But rumors abound.

For example, since August 2021, Wired reports that industry experts predict that GPT-4 will contain 100 trillion parameters. Building an AI with more parameters does not necessarily guarantee better performance; However, it can affect the response.

Other rumors point to better computer code generation and the ability to generate images and text from the same chat interface. An artificial intelligence that can create a video is also expected.

Working with text, images, and video is known as a multimedia form. Machine learning expert Emil Wallner chirp that GPT-4 may have this capability.

Takeaway:

– GPT-4 is coming, but improving Codex is the focus
– The next edition of the Codex Alimentarius will be important
– GPT-4 will have a similar but better trained model size
– DALL-E will become public
– They bet that multimedia forms will outperform pure text forms when creating text https://t.co/sDQ3mvCvuI

And[مدش]. Emil Wallner September 6, 2021

A hint of potential GPT-4 specifics came from a price sheet shared by Travis Fisher on Twitter. A new developer product called Foundry will include more than one model, adding two “DV” models that can handle more words at once as input (maximum context).

OpenAI notably announced a new developer product called Foundry, which enables customers to run OpenAI model inference at scale with custom capacity.

It also reveals that DV (Davinci; likely GPT-4) will have up to a 32KB maximum context length in the public release. 🔥 pic.twitter.com/5KEsWLqPdc

And[مدش]. Travis Fisher (transitive_bs) February 21, 2023

GPT-3.5 works with up to 2048 words, but the two DV models, which may be GPT-4, have four times (8K resolution) and 16 times (32K) that capacity, allowing for deeper discussions to solve much larger challenges.

In the end, this is all just speculation, and all we know for sure is that GPT-4 is a work in progress and that it could significantly improve the outcomes possible with ChatGPT.

Is there a release date for GPT-4?

Laptop opened on ChatGPT site.
stock struggle

GPT-4 is coming, and it will probably arrive this year. However, we don’t have a confirmed release date yet when it will be released in ChatGPT. the The New York Times suggested that it will arrive as soon as the first quarter of this year. Since we’re in March, that will launch in a matter of weeks.

Microsoft has confirmed that GPT-4 will arrive the week of March 13th, and that it will be a multimedia model that can create text, images, and even videos.

However, it is not clear now how Microsoft and OpenAI will benefit from GPT-4. The form will likely launch for search, so it may be a little longer before we see it in Microsoft’s ChatGPT and Bing Chat.

Does Bing Chat use GPT-4?

A preview of the new Bing chat can be seen even on a MacBook.
Photo by Alan Trulli

Microsoft states that the new Bing file or Bing Chat is more powerful than ChatGPT. Since OpenAI Chat uses GPT-3.5, there is an implication that Bing Chat can use GPT-4. This has not been confirmed.

Obviously, Bing Chat has been upgraded with the ability to access current information online, which is a huge improvement over ChatGPT, which can only benefit from the training it has received until 2021.

In addition to accessing the Internet, the AI ​​model used in Bing Chat is much faster, which is very important when you take it out of the lab and add it to a search engine.

It seems unlikely that it would be equivalent to OpenAI’s GPT-4 model. There would be no need to be careful if GPT-4 was already publicly available.

GPT-4: Evolution, Not Revolution

While GPT-4 would certainly be impressive, OpenAI CEO Sam Altman said in an interview published by StrictlyVC: Connie Loizos on YouTube states that “People beg to be disappointed, and they will be disappointed.”

StrictlyVC in Conversation with Sam Altman, Part Two (OpenAI)

Altman acknowledges the potential for artificial general intelligence to wreak havoc on world economies, and has expressed that a rapid deployment of many small changes is better than a shocking advance that provides little chance for the world to adapt to the changes.

In short, GPT-4 will be an evolution, not a revolution. According to Altman, the next version of ChatGPT will not be AGI and will not contain 100 trillion parameters. These rumors are not true.

Editors’ recommendations






Share:

Leave a Reply

Your email address will not be published. Required fields are marked *